in different latitudes. On the Winipeg it was accompanied by the purple swift (Progne purpurea), whose northern limit we did not ascertain.

We resumed our voyage at three in the morning on the 1st of August, and when we landed to cook breakfast, saw some recent footmarks of Eskimos. As these people are employed at this time of the year, in hunting the rein-deer on the hills which we were skirting, we were in constant expectation of seeing some of their parties. The Rein-deer Hills, as viewed from the eastern channel, seem to be an even-backed range; but when examined with the telescope, they are seen to consist of many small, oblong, rocky eminences, apparently of limestone, and are sparingly wooded. In the course of the morning we crossed the mouths of three pretty large affluents, coming in from the hills, and also two cross canals, dividing M'Gillivray Island into three sections.

About thirty-five miles from Point Separation, or in latitude 68° 10′ N., the channel washes the foot of a low dome-shaped bluff, in which the intrusion

TRAP

SANDSTONE

LIME

STONE

of a mass of trap, which now forms the top of the hill, has tilted up a bed of limestone, and separated it from one of sandstone.

In the afternoon we passed another considerable affluent from the hills in lat. 68° 18′ N.; some hours later, another one of less size; and very soon afterwards crossed a channel which bounds Harrison Island on the south. This island, like M'Gillivray's, is divided into several portions by minor creeks. The boats were under sail all the afternoon, and must have been observed, about 5 o'clock, by the hunting-parties of the Eskimos, for at that time we noticed a line of six or eight signal smokes, raised in succession along the hills, and speedily extinguished again. As the Eskimos use fire-wood very sparingly for cooking, and, like the Indians generally, burn only dry wood which emits but little vapour, we knew that the smokes we saw were intended to spread the intelligence of the arrival of strangers in the country, and therefore that we might expect to find a considerable body assembled on some part of the river to meet us. In the evening we landed to cook supper, and re-embarked to continue under sail all night, with a very light breeze; our progress was, however, slow, owing to the uncertain eddies and currents, produced by the junction of the several cross-channels. At midnight we passed the creek which bounds Harrison Island on the north, in 68° 37' N. Here several

gently swelling elevations interpose between the river and the main ridge of the Rein-deer Hills. The valleys and borders of the river are well wooded, but the summits of the eminences present only scattered spruce firs, with stunted tops and widely spreading depressed lower branches. The canoebirch (Betula papyracea) is frequent, and the trees we measured were about five inches in diameter. The Populus balsamifera and Alnus viridis grow to the height of twenty feet, and the Salix speciosa to upwards of twelve. The Ribes rubrum, Rubus chamamorus, and Vaccinium vitis idea, bore at this time ripe fruit. The Rosa blanda, Kalmia glauca, Nardosmia palmata, and Lupinus perennis, were also observed flourishing in this high latitude, together with several other plants which extend to the sea-coast. Among the birds, we saw the great tern (Sterna cayana), the Coryle alcyon, and Scolecophagus ferrugineus, the latter in flocks.

August 2d. For five or six hours this morning we ran past the ends of successive ridges separated by narrow valleys. The diagram gives the outlines of one of these spurs seen on the southern flank. It is about three hundred feet high, and its acclivities are furrowed deeply, producing conical eminences which are impressed with minor furrows. The vegetation is scanty; a few small white spruces straggle up the sides; and the soil, where it is

exposed to view, is a fine white sand. Large boulders lie on the sides of the hills, and, judging from the structure of the only point on which time permitted me to land, the whole appears to be similar to the sand deposit with its capping of boulder gravel which covers the shale on Athabasca and Mackenzie Rivers. On the point in question, the white sandy soil was ascertained to come from the disintegration of a sandstone, which has just coherence enough when in situ to form a perpendicular bank, but crumbles on being handled. It consists of quartz of various colours, with grains of Lydian-stone loosely aggregated, and having the interstices filled with a powdery matter, like the deposits of some calcareous springs. Similar sandstones occur at the "Narrows." Above it, there is a bed of gravel, also formed of variously coloured

grains of quartz, mixed with chert from limestone. Most of the quartz is opaque, and veined or banded, but some of it is translucent. Some bits are bluish, others black, and many pebbles are coloured of various shades of mountain green. The latter are collected by the Eskimos and worn by them as labrets. The gravel covers the whole slope of the point, which is so steep as to require to be ascended on all fours. In one part a torrent had made a section of a bed of fine brown sand, twenty feet deep. On this bank I gathered the Bupleurum ranunculoides, which grows in Beering's Straits, but had not been found so far westward as the Mackenzie before; also the Seseli divaricatum, which had not been previously collected to the north of the Saskatchewan.

In latitude 68° 55′ N. the trees disappeared so suddenly, that I could not but attribute their cessation to the influence of the sea-air. Beyond this line a few stunted spruces, only, were seen struggling for existence, and some scrubby canoe-birches, clinging to the bases of the hills. Further on, the Rein-deer Hills lowered rapidly, and we soon afterwards came to Sacred Island, which, with the islets beyond it, is evidently a continuation of the sandy deposit noticed above. Had time permitted, I should have gone past Sacred Island, northwards, to deposit some pemican on Whale Island, but at so advanced a period of the summer,